© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 66724
    Prev
    Next
    INDEX
    Explanations

    talks about

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron fires on verbs and nouns that signal investigation or explanation—words like “probe,” “dissection,” “look,” “talks,” and “discusses” that introduce analysis or explanatory content.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    <bos>
    -1.79
    gencias
    -1.71
    in
    -1.38
    pinephrine
    -1.37
     incredibly
    -1.36
    bende
    -1.34
     meinem
    -1.30
    stwa
    -1.28
     greatest
    -1.27
     insanely
    -1.27
    POSITIVE LOGITS
     how
    3.45
     what
    1.74
    ſſ
    1.57
    ௅
    1.49
    どう
    1.48
    どの
    1.39
    屨
    1.38
    (^_^;)
    1.36
    Curiosidades
    1.34
    izmus
    1.34
    Activations Density 0.091%

    No Known Activations